In this project I have tried to understand understand all the different decision boundries made by different AI elementary model. In this project we see how kernels can be used with SVM to give non linear functions. First we convert our iris data 4 features into two features for this different methods have been used
- Choosing the feature based on PDF of all the features.
- Choosing the feature based on PCA
- Choosing the feature based on T-SNEE After convering our feature in 2D we train the data and plot decision surface for
- KNN
- Naive Bayes
- Logistic Regression
- SVM a) Linear SVM b) RBF Kernel c) Polynomial Kernel
- Decision Tree
The framework used in this project is sklearn You can run the program in Google Colab